Washington, DC – On a press call and webinar held today, Matt Barreto, UCLA Professor of Political Science and Chicano Studies, and Co-Founder of Latino Decisions, presented the results of the Latino Decisions Election Eve poll. The key finding: Latinos backed Hillary Clinton over Donald Trump by a 79-18% margin.
Barreto also highlighted why the national exit polls miss the mark when it comes to capturing sub-groups such as Latino voters, and why the exit poll estimate that Clinton beat Trump by a narrower 65-29% margin among Latinos should be greeted with extreme skepticism.
As pointed out in the presentation, the Latino Decisions finding on national presidential margin is consistent with high-quality, large-sample, bilingual polls carried out by a number of groups in recent months. For example, the Latino Decisions’ Election Eve finding is similar to the findings of theUnivision/Washington Post poll, which had Trump at 19% among Latinos; the NBC/Telemundooversample which found Trump at 17%; the NALEO/Telemundo poll which found Trump at 14%; the FIU/New Latino Voice which found Trump at 13%; and the Justin Gross statistical model that found Trump at 18%.
Further, Barreto pointed to an examination of the actual election results from counties and precincts which are majority Latino show higher rates of Latino voter turnout in 2016, and show Clinton winning roughly 80% of the Latino vote. Barreto also presented initial findings assessing Latino turnout across the country. In county-level analyses of Latino-heavy sections of Florida, New Mexico, and Texas, for example, Latino voters were voting in higher numbers and in candidate margins similar to the margin found in the Latino Decisions Election Eve poll.
During the presentation today, Barreto pointed out that by its own admission, Edison Media Research, who carries out the exit polls, admitted that its sampling “is not designed to yield very reliable estimates of the characteristics of small, geographically clustered demographic groups.” He noted the lack of transparency in the selection of precincts to sample, and suggested that journalists should ask the following questions of the exit polls before accepting them as useful for conclusions regarding Latino voters: “Which precincts did they select? How many were in Latino neighborhoods? How many Spanish interviews did they conduct? And did they match Latino sample to known Census demographics?”
Click here to view Barreto’s slide presentation and here for more information on what constitutes a good versus bad Latino poll
Click here for national and state toplines and crosstabs for the Latino Decisions election eve poll
The full toplines and crosstabs for the national and state-specific Latino Decisions Election Eve poll, sponsored by America’s Voice, SEIU, Mi Familia Vota, NCLR and Latino Victory Project, can be found here.
Audio from today’s call is available here.
About the Latino Decisions Election Eve Poll Methodology
Latino Decisions completed 5,600 interviews with Latinos with strong turnout histories who reported having already voted, or reported being certain to vote in the November 8, 2016 presidential election. Interviews were conducted either online or via telephone with live callers, all of whom were bilingual, and both phone and web interviews were completed in the language of preference of the respondent. Respondents were reached on landline (40%) and cell phones (35%), or invited to complete the survey on the web (25%), between November 2-7, 2016 and interviews averaged 10 minutes in length. The results are weighted to match national Census demographics for Latinos, including national origin groups, age, education, place of birth, and gender.
The national sample (N=5,600) carries an overall margin of error of +/-1.8 percentage points. Florida has 800 completed interviews and carries a margin of error of +/-3.5 percentage points. The remaining individual states sampled — Arizona, California, Colorado, Illinois, New York, Nevada, North Carolina Ohio, Texas, Wisconsin, and Virginia – have 400 completed interviews and carry a margin of error of +/-4.9 percentage points. For more detail, see 2016-election-eve-poll-methodology
####
